It is a barometer showing that performance is indeed impacted.
Think much?
The screensaver isn't an accurate barometer of its ability to render websites, or play video, or efficiently handle memory and paging. Real world usage (which is what you kept railing against with your screensaver test) matters much more than your decided benchmark. Does the thing do what I need it to do and does it do it efficiently? If the answer is "yes" then it works. I don't use a screensaver so your metric is useless to me.